This full text version, available on TeesRep, is the post-print (final version prior to publication) of:
Ryan, C. G. et. al. (2011) 'Sitting patterns at work: objective measurement of adherence to current recommendations', Ergonomics, 54 (6), pp.531-538. For details regarding the final published version please click on the following DOI link: http://dx.doi.org/10.1080/00140139.2011.570458 When citing this source, please use the final published version as above.
This document was downloaded from http://tees.openrepository.com/tees/handle/10149/129838 Please do not use this version for citation purposes. All items in TeesRep are protected by copyright, with all rights reserved, unless otherwise indicated. TeesRep: Teesside University's Research Repository http://tees.openrepository.com/tees/
Title Sitting patterns at work: objective measurement of adherence to current recommendations. Running title Sitting patterns at work. Authors names Cormac G Ryan1, P Margaret Grant, Philippa M Dall, Malcolm H Granat. Affiliation School of Health, Glasgow Caledonian University, Scotland, UK, G4 0BA.
Corresponding Author: Dr. Cormac Ryan Ph.D. telephone (0044) (0)1642 384916; email:
[email protected] 1
1
Abstract Long uninterrupted sedentary periods, independent of total sedentary time, are risk factors for poor health. There is little objective data relating to workplace sedentary behavior and adherence to current recommendations. The sitting behaviour of office workers (n=83) was quantified objectively using body-worn accelerometers (activPAL™) over a working week. Adherence to three different recommendations (maximum length of a sitting event of: 20minutes, 30minutes and 55minutes) were assessed. Participants were seated at work for 5.3±1.0hours/day (Mean±1SD), equivalent to 66±12% of the working day, accrued in 27±7events/day individual sitting events. Dependent on the recommendation applied, 5-20% of sitting events and 25-67% of time was accumulated in sitting events longer than current guidelines. No participants met the 20 or 30minutes recommendations on every working day but seven (8%) participants met the 55minutes recommendation. In conclusion, office workers spend a considerable period of their day sitting, accumulated in uninterrupted sitting events longer than current recommendations. Key words Sedentary; sitting; behaviour; adherence; activPAL™
Statement of relevance Emerging evidence suggests prolonged sitting has negative health effects. In this study of office-workers 25-67% of time sitting was accumulated in events longer than minimum recommended durations. Adverse sitting behaviour is prevalent in the office making it an appropriate setting to target the reduction of this behaviour.
2
1. Introduction Moderate to vigorous physical activity (MVPA) can have a beneficial effect on many health conditions such as cardiovascular disease, type II diabetes and obesity (Pate et al. 1995; Haskell et al. 2007). Very specific MVPA guidelines for adults exist, recommending 30 minutes of moderate intensity activity on most days of the week to be accumulated in bouts of 10 minutes or more (Haskell et al. 2007). It has been argued that many modern day occupations require too little MVPA and that strategies should be put in place to increase physical workloads in the workplace (Straker and Mathiassen 2009). Sedentary behaviour can be defined as time spent sitting (Biddle 2007; Katzmarzyk et al. 2009; Owen et al. 2009) or sitting/supine (Chastin and Granat 2010; Jans et al. 2007). Recently, it has been shown that sedentary behaviour is a risk factor for many health conditions often independent of MVPA (Healy et al. 2008a; Katzmarzyk et al. 2009). Additionally, there is preliminary evidence that it is not just the total sedentary time which is relevant but the manner in which it is accumulated; longer sedentary events being associated with undesirable health markers (e.g. a large waist circumference) compared to shorter sedentary events (Healy et al. 2008b; Healy et al. 2011). Despite this, there are no official national guidelines for sedentary behaviour. The recommendations that do exist are few in number, vary in their advice, and appear to be based upon expert consensus rather than robust scientific evidence (Atlas and Deyo 2001; Chartered Society of Physiotherapy 2005; Owen et al. 2009). Many occupations are computer-based, and this results in predominantly seated behaviour at work (Hill and Peters 1998; Smith et al. 1999). Data from a range of industrialised countries, including Sweden, U.S.A, Australia and New Zealand, suggests that increased sedentariness at work is an international phenomenon (Straker and Mathiassen 2009). While the biomechanical aspects of seating and sitting have been subject to extensive research (Corlett 2006; Page et al. 2009), behavioural aspects of sitting at work have not received the same level of attention. McCrady and Levine (2009) identified that workers in the USA were sedentary for approximately two hours more on working days compared to leisure days. Workers were classified as having jobs that were either “entirely sedentary” (chair bound most of the day) or “semi-sedentary” (intermittently standing and chair-bound but without substantial 3
walking or physical labour) (McCrady and Levine 2009). A recent report on the sitting behaviour of office-based workers in Australia found employees were sitting for 77% of their time at work (Thorpe et al. 2008). Thus the work place is a setting where sedentary behaviour is highly prevalent, and this presents an important location where strategies to reduce sedentary behaviour should be introduced (Hamilton et al. 2008; McCrady and Levine 2009). However, without clear evidence based guidelines targeted strategies are difficult to design and implement. A literature search identified three different recommendations which specified maximum sitting time; 20 minutes (Chartered Society of Physiotherapy 2005), 30 minutes (Atlas and Deyo 2001), and a five minute break every hour (Owen et al. 2009). The 20 minute recommendation was taken from the Chartered Society for Physiotherapy, Fit-for-work initiative (Chartered Society of Physiotherapy 2005) and the 30 minute recommendation was based upon information given to individuals with acute low back pain (Atlas and Deyo 2001). The development of both of these recommendations was rooted in musculoskeletal medicine. In contrast, the five minute breaks every hour recommendation, was developed with a focus on metabolic health, and was based upon the potential clinical implications of normative objectively measured sitting data (Owen et al. 2009). However, all three of these recommendations appear to be based upon authors’ professional judgment rather than robust scientific rationale. A first step towards developing more evidence based guidelines should be to collect objective normative data of sitting behaviour in the work place and assess adherence to current recommendations. Whilst recent research has objectively quantified total sitting time at work (Thorpe et al. 2008; McCrady and Levine 2009), this data has been collected using accelerometer counts as a proxy for sitting times rather than using devices that directly measure sitting. Additionally, there has been no published data on sitting patterns such as the length of individual sitting events. Such information may help develop an understanding of what type of sitting recommendations may be practical to achieve and guide future interventions to reduce sedentary behaviour. The aim of this study was to objectively quantify patterns of sitting behaviour in an office setting and to assess adherence to current sitting recommendations
4
2. Materials and methods 2.1 Design In this observational study, which is a reanalysis of data collected for another study, the sitting behaviour of office-based employees during working hours over a oneweek period was recorded using activity monitoring. The total sitting time accumulated and the patterns of sitting behaviour were quantified and compared to current sitting recommendations within the literature. 2.2 Participants Participants for this study were selected from a database of free-living physical activity profiles of healthy adults. Participants were recruited onto the data-base using recruitment emails circulated to staff and postgraduate students of Glasgow Caledonian University. Participants were included on the data base if they were aged ≥18 years, lived in the community, had no mobility problems and were not expecting to experience any unusual activity in the monitoring week (e.g. on vacation). Participants from the database were selected if they worked in an office-based occupation, classified as entirely sedentary (chair bound most of the day) or semisedentary (intermittently standing and chair-bound but without substantial walking or physical labour) as defined by McCrady and Levine (2009). For individuals that did not provide clear indication of which days were working and non-working, weekdays were considered to be working days. Any individual who worked part-time but did not indicate which days they worked were excluded from the data set. Participants were also excluded from the data set if the number of working days recorded was less than three days. Ethical approval was received from the Glasgow Caledonian School of Health and Social Care research and ethics committee. All participants provided written informed consent. 2.3 Procedure Sedentary activity was measured over a seven day period between October 2007 and October 2009 in Glasgow, Scotland. A time frame of 9:00am to 5:00pm on working days was selected to represent time at work. This time frame was selected to represent the usual working hours of Glasgow Caledonian University staff. Sitting events which
5
crossed this time boundary were also included if more than half of that event was within the 9:00-5:00pm time frame. 2.4 Physical activity monitor Physical activity was measured using the activPAL™ physical activity monitor. The activPAL™ is an accelerometer based device which is worn on the thigh. The monitor produces a signal which is related to the inclination and movement of the thigh. From this signal it can directly identify time spent in the postures of sitting/lying, standing and walking on a second by second basis. The monitor has been shown to be a valid (>99% sensitivity and specificity) and reliable (ICC >0.99) measure of sitting posture (Grant et al. 2006). The activPAL™ is small and unobtrusive requiring little input from the wearer. These issues of accuracy and low burden are important when measuring exposure to risk factors in an office environment (Van Eerd et al. 2009). 2.5 Data analysis Three sitting activity outcome measures were calculated for the whole group and for occupational subgroups; 1) the total number of sitting events/day 2) the total times sitting/day and 3) the mean of the longest sitting periods for each participant. Adherence to sedentary recommendations was then assessed. Adherence to three separate recommendations were used in this study; a maximum sitting time of 20 minutes (Chartered Society of Physiotherapy, 2005), a maximum sitting time of 30 minutes (Atlas and Deyo, 2001), and a five minute break every hour (Owen et al. 2009). The latter recommendation was operationalised as a maximum sitting time of 55 minutes. The activPAL™ software (version: 5.9.1.1) classified the acceleration signal into sit/lie, stand and walk. In-house purpose built software (written in visual basic for applications, to run in excel), was used to present data by event and to select data between 9 and 5pm on work days. Visual inspection of the day was used to identify workdays and check whether the first and last periods should be included. A number of different outcome measures were calculated from the sitting data. For the whole sample, total sitting time, total number of events and total time and number of events longer than 20 minutes, 30 minutes and 55 minutes were derived. An event is defined 6
as a single uninterrupted sitting period. These were calculated from cumulative frequency graphs of sitting activity as a proportion of number of events and total time. For each participant, the longest sitting event was reported. In addition, the number of sitting events greater than 20, 30, and 55 minutes in each day was calculated. From this, the number of days in which the sitting recommendations were met for each individual was identified. Individuals were deemed to have met the guidelines if they met the criteria for each working day assessed. Different subgroups of office-based worker, based on self-reported job description, e.g. lecturer, technician were identified. Differences between subgroups for basic demographics and the key sitting outcome measures were compared using a one-way ANOVA with the Statistical Package SPSS (Version 18.0). 3. Results From a physical activity database of 121 individuals, data from 83 participants were included in this study. The remaining 38 participants were excluded as; they were not office-based, there was insufficient information to allow classification as an officebased employee or to identify their work days, or there were less than three working days recorded. Of the 83 participants included, four occupational subgroups were identified; 27 lecturers, 27 researchers (desk-based), 7 technicians and 22 administrators. One way ANOVA revealed significant differences between occupational subgroups for age (lecturers were statistically older than researchers and administrators) and BMI (no significant post-hoc differences); there were no statistically significant differences in descriptive characteristics between the subgroups. Participant characteristics are shown in Table 1. Data were available over five working days for the majority of participants (n=60). The minimum number of working days recorded was three (n=6) and the maximum was six (n=1). Insert Table 1 here 3.1 Sitting time For the group as a whole, 66±12% [range 31-90%] (mean±1SD [range]) of time at work was spent sitting. The total time spent sitting by participants during 388 working 7
days was 2,042hours (5.3±1.0 hours/working day [range 2.5-7.2]), which was achieved in a total of 9,918 (27±7 events/working day [range 13-48]) individual sitting events. The mean of the longest sitting event of each participant was 98±34 minutes [range 43- 201]. There was no statistically significant difference between the four occupational subgroups for total sitting time or the number of sitting events, though there was a trend for the technicians to spend less time sitting and accumulate more sitting events (Figure 1). There was a statistically significant difference between the occupational subgroups for the mean longest sitting period (p
